ADP down while BLS up. Aggregate hours down.
Figure 1: Manufacturing employment from BLS CES (blue, left log scale), from ADP (red, left log scale), in 000’s, s.a., total hours in manufacturing, production and nonsupervisory workers, in logs 2025M01=0 (teal, right scale), s.a. Source: BLS, ADP via FRED, and author’s calculations.
Off topic – Secretary Hegseth has fired Secretary of the Army Driscoll and Army Chief of Staff George. The BBC is reporting that Hegseth fired both men because he fears for his job and sees Driscoll, in particular, as a rival.
There were already rumors that the war-criminal-in-chief is cuing Hegseth up to take the blame for the war on Iran, and these firings suggest Hegseth has heard the rumors. In a well-run administration, decapitating one’s own Army during war time to protect one’s own career would end that career. In this administration, Hegseth, rube that he is, has handed the war criminal an easy way to throw him under the bus for the Iran war. Driscoll would have reason to feel loyalty to the war criminal if he were brought back in as Secretary of Defense.
i usually don’t put much stake in the household survey with it’s large margin of error, but it seems to be trying to tell us something….the labor force participation rate at 61.9% in March is a 52 month low; those not in the labor force rose by 488,000 to another record at 104,771,000, the employment to population ratio was 59.2% in March, a 53 month low, and their seasonally adjusted employment figure of 162,848,000 is down by 1,144,000 since December and down 661,000 year over year..
https://fred.stlouisfed.org/series/CE16OV
Some of the decline in the employment-to-population ratio is probably due to the aging of the workforce. Here’s a comparison of the prime-age participation rate to the overall participation rate:
https://fred.stlouisfed.org/graph/?g=1UrU4
Notice that prime-age participation is near the top of its historic range, while the overall rate has been declining for some time. That’s baby boomers aging oit ofthe wotkforce. I suspect the slow-hire, slow-fire performance of the labor market is hurting participation among older workers, but that’s just a guess.
i just remembered that when they do the annual population revisions to the household survey, they only revise the current and future months, and leave the old data as it was…that accounts for the big drop in all the Current Population Survey data in January, and likely at least some of the other longer term anomalies that i cited..
rjs: I have taken to using by default the experimental BLS series using smoothed population controls. See graph here.